I have two arrays right next to each other and I have noticed that one goes into resting in the afternoon while the other is working. What would cause one array to go resting in the afternoon while its partner is working fine. Neither arrays are shaded in the afternoon. In general, one array generates more power than the other. The array on the right outperforms the array on the left.
Are these Classics ? Kids ? Both or even other controllers are similar...
Looks to me like the one on the top graph went into Absorb and was doing the work at that time. That is not a problem in itself but usually is because the battery voltage on the top one is reading slightly lower than the one on the bottom.
When the two controllers are in Absorb or FLoat, it is pretty common for one to do that.
Look at the display on each Classic and see how far off they are. If they are off by a lot, say 1 volt or more, then you might
be able to go into TWEAKS and adjust them a bit more towards each other.
Another thing that might happen is if each Classic thinks that the battery temperature is different than the other Classic.
Are the two controllers tied together like in Follow Me mode ? If so, then the battery temperature can be tied together too.
If that bottom controller really isn't able to pull its weight and the 2 PV arrays are very similar, then I would look at the PV input voltages while they are in "MPPT" mode (bulk, float mppt) and make sure they are close in V. This would check the tracking of the controller.
They can be different if there is partial shading but should typically be running at around 80% of the open circuit voltage (Voc) of the PV array when tracking without shading.
You might have to turn off the other controller if it is doing all the work and keeping the controller in question from running..
If it IS a tracking problem and if it is a Classic, try changing the mode from SOLAR to Legacy P&O and see if that helps.
Yep, two Classic 150s. Good point about the voltage and the battery temp. They are not using follow me but that would be really easy. I have the battery temp sensors on different batteries and I think I can see the sensed temp somewhere. I can remote VPN into my cabin( I had internet before I had water) and check some of the settings and the voltage with the local app.
I zoomed in a little to highlite the difference.
local app has both classics showing battery voltage at 25.2, batt temp is close.
Local app shows double the output.
OK, so all is close. The top LA unit's battery temperature is only 1 degree C less than the bottom so at 24V, I think that turns out to be -5mV X 12 = 60 millivolts difference. Less than 0.1V so not much to make a difference, really.
Check the PV input voltage during the day when the units are in MPPT mode and take a note of that referenced to Voc.
That percentage should give you a good idea if they are tracking the same. If they typically run at about the same voltage or percent of Voc, then they are tracking properly and grabbing as much power from the array as they can.
If those MPPT Vmp / Voc is something way different than 0.80 then try Legacy P&O mode.
I neglected to invlude the panel info, they are the mission solar 315
I added PV volts to the mymidnite display from today and created screen shots for the same time period. Flaky 114.4 vs solid 105.9. Thats the best I can do until get up there to measure directly in the panels.
Quote from: schroew on December 27, 2019, 12:04:19 AM
I added PV volts to the mymidnite display from today and created screen shots for the same time period. Flaky 114.4 vs solid 105.9. Thats the best I can do until get up there to measure directly in the panels.
OK, this is a good start. SO your Voc is around 114V then it would appear.
Can't tel a lot from the 105V in Absorb but when the Voc is like this, say 114V, in Bulk MPPT the input voltage should be somewhere around 90 volts or so... give or take just a few volts. Maybe 75V to 95V or so in MPPT mode
.
While I was there today I noticed that the performing controller was showing steady watts and volts while the flaky one was continually jumping around WRT the watts and the volts.
Quote from: schroew on December 29, 2019, 05:16:30 PM
While I was there today I noticed that the performing controller was showing steady watts and volts while the flaky one was continually jumping around WRT the watts and the volts.
Assume that the Volts that were "jumping" around was the PV in, and not the battery voltage, read on that Classic.
In any event, please do check ALL DC connections. Screw terminals on bussbars, breakers and Classic DC connections on the blue terminal strip can loosen, in time. One really should check all of these, 24-48 hours after initial torqueing.
Also check all of the PV cable connectors, and connections.
As you know, never break or make a connection of the MC PV cable connections, when there is any current flowing through those leads.
Just a thought, good luck, Vic
video-
Quote from: schroew on December 29, 2019, 06:34:15 PM
video-
Excellent ! Video is worth several words....
That jumping around, which isn't really changing much, would be normal, especially with two Classics in the system.
It has to move the input voltage around because the Classic is in voltage regulation mode. The battery is sitting pretty steady at 29.2 volts and to do that, the PV input voltage has to move around with changing conditions and load. AND the other Classic, if also in Absorb, is contributing to the voltage regulation.
Changing PV voltage is how voltage regulation is done in MPPT charge controllers.
It was in absorb. Screenshot attached. I will check the connections on the controller ( I checked the combiner last weekend). Maybe if everything is tightly connected this is normal. Would I benefit from setting up follow me so the charge controllers can coordinate their efforts? To be fair I just got this running in late November so I don't have alot of data points but from whst I hsv seen so far the behavior is consistent. Does this look at all like a panel issue?
If it matters the array thats flaky was the first one setup by a couple of days.
Quote from: schroew on December 29, 2019, 10:51:38 PM
If it matters the array thats flaky was the first one setup by a couple of days.
Follow me will help both controllers to be in bulk / absorb at the same time and go to float at the same time.
What was the difference in time of day between those two screen shots ? The newer still and the earlier video ?
The screen shot is from the video. I just did a screen shot because I fumbled the phone hitting the status button to wake up the display on the controller.
To add to boB's post, Follow me will NOT help keep one controller from going to resting when the battery is full. Essentially the multiple classics all see a very slightly different battery voltage and as the needed power to maintain the full level drops and the controller's arent needed the ones that see the battery as the highest stop charging.
I wrote a logstash config for the data exported from mymidnite and when you look at the percent VoC ( bob suggestion), power, PV in kibana timelion they don't look that far apart. Who has a mapping for some of the fields like mppt_mode, resting, charge_stage, info_flags*?
Quote from: schroew on December 31, 2019, 01:14:25 PM
I wrote a logstash config for the data exported from mymidnite and when you look at the percent VoC ( bob suggestion), power, PV in kibana timelion they don't look that far apart. Who has a mapping for some of the fields like mppt_mode, resting, charge_stage, info_flags*?
I'm not sure if this list maps directly to My MidNite data but here is what the Classic/MNGP uses for menu status from
modbus register address 4119 (register 4120) that is a 16 bit number divided up into two 8 bit parts.
These numbers are in the high 8 bit half I believe. The low 8 bits are just a state machine number...
I don't think I was able to paste this as courier font, sorry if it formats awful...
#define StatDispResting 0 //Resting
#define StatDispAbsorb 3 //Absorb
#define StatDispBulkMppt 4 //Bulk MPPT
#define StatDispFloat 5 //Float
#define StatDispFloatMppt 6 //Float MPPT
#define StatDispEqualize 7 //EQ
#define StatDispPvShade 8 //Do not need this state -- Show as Info
#define StatDispOverLd 9 //Do not need this state -- Show as Info
#define PvHyperVoc 10 //PV over Volt //Do not need this state -- Show as Info
#define WakeupCountDwn 11 //Classic waking up (power up only... Wait)
#define ArcFault 12 //Show as Info
#define GroundFault 13 //Show as Info
#define DefCon4Err 14 //ERROR (Defcon 4) (unused now ?)
#define ReCal_NegCurr 15 //Negative Battery or PV Current (not snow melt)
#define CurrentLimitReached 16 //Sliding current or user current limit touched
#define UnitIsHOT 17 //could also invoke sliding current limit
#define StatDispEqMppt 18 //EQ MPPT... Trying to get to EQ voltage
Ahhh ! Looks like it DID format correctly using the [courier] tag ! Great !
hmm, let me see how t works.
one record from the rubydebug output in logstash.
Quote from: schroew on December 31, 2019, 09:47:36 PM
one record from the rubydebug output in logstash.
Ahhh... Nice !
So, the matching number is that third line down, the stage which in your picture is sitting at 0 or Resting